Achieving Efficiency Through Robotic Process Automation

INTRO

On this episode of the InsurTech Geek Podcast, we’re talking about Achieving Efficiency Through Robotic Process Automation with Dan Savage from MassMutual.

The Insurtech Geek Podcast is all about technology that’s transforming and disrupting the insurance world. We’ll be interviewing guests and doing deep dives in the specific tech we see changing the industry. We’re taking you on a journey through insurance tech, so enjoy the ride and geek out!

INTERVIEW

JAMES: And we are back with another great episode of the InsurTech Geek Podcast. I’ve got my main man, my fellow Texagander, my golf player extraordinaire, my most interesting man in insurance. That’s right, he’s with us. Rob Galbraith, how’s it going, Rob?

ROB: It’s going James, it’s going. Started to pick up the game again after a couple of decades away from it. My college age son, he’s finally belatedly got hooked on golf. So I’ve been going out with him and, I still can hit it past him, you know, three four yards, and this is the only guy field end.

JAMES: Did you lay a little beat down on your son?

ROB: Little beat down. I mean the scoring isn’t close, but who cares about scoring like that? That’s it’s more important about the drive, so still pounding it by him. He’s a lot taller. He’s a lot more muscular. I wish I had his genes. I don’t know where they came from.

JAMES: Well, that’s great, man. I mean it. Look out, driving your college age son is every dad’s dream, you know, got to put them in his place. Remind him that the old man still got it.

ROB: I’m going to be eligible for the senior tour here, that are the champions here. They call it now in a month or so, I’m not ready for that yet.

JAMES: I’ll tell you what, I just picked up my daughter from Boston, from Baston. She was at Boston Conservatory for three weeks, she’s 17, gonna be a high school senior at Interlochen Arts Academy, but she went to Boston Conservatory for this summer camp for musical theater dance and watching her perform, it’s just so incredible. Like she so far exceeds any place I ever got to is singing or dancing. And because I’m in the piano guitar singing, I’m in a ballroom dance team, and she blows anything away that I’ve ever done musically. So it’s exciting, I mean, it’s really exciting to watch your kids succeed. But it’s also exciting to put them in their place sometime and show them. Hey, look, you know, I still got some skill. I still got some skill. That’s great, I’m glad to hear it went well. And obviously good to see you.

And with us today, our special guest, Dan Savage from MassMutual, joining us from beautiful Connecticut. Dan, how’s it going?

DAN: Great, James, thanks for having me on, appreciate it.

JAMES: Man, glad to have you on. It’s always great to hang out with fellow InsurTech geeks who are putting technology into action every day in real insurance companies. You know, no theoretical stuff here. We’re talking about real deal, real deal technology implementation, which is what I like hearing about.

And before we get into talking about what you’re doing at MassMutual about robotic process automation, about all those fun things, you can really radically transform organizations, one little automated process at a time, we’re gonna talk about you. You’ve got a really neat local story, right? You are about as local as it gets there in Connect I Cut. That’s, of course, how I learned growing up in Louisiana. They had to keep it simple for me, and so that’s how they taught me how to spell your state. So you were born and raised in Connecticut, near the Massachusetts. Then you went to college in Connecticut and then you went to work at MassMutual in Connecticut. And so you’re still in that same radius when you see your high school friends, your college friends, and your work friends in the same night.

DAN: I wouldn’t say that now that I’ve got two kids, but yeah, I’ve been in this area for my whole life. Like I said, I grew up kind of closer to the mass border. I work up in Springfield now. But you know, MassMutual has had offices in Connecticut as well, I worked in some of those. I’ve been here my whole life and you know if you’re gonna get into insurance, this is a decent place to get into it. The Hartford’s, the insurance capital of the world at least that’s what they say around here.

JAMES: Yeah, maybe in America, we’ll hold back on the world.

DAN: There you go, and MassMutual is right up the road ah about 25-30 minutes from there.

JAMES: Oh, wow, that’s great. And, and what led you, I mean, you know, you didn’t, like when you kind of look back, when you were a kid growing up in Connecticut, like what’d you want to be when you were a kid career wise? Like I wanted to be an astronaut. What would you, what did you aspire to be?

DAN: I guess it depends on what age, right? So like when you’re a little, kid little, little kid, I don’t remember what I wanted to be. As I was growing up, I got super into soccer. That was a big thing. My dad played college soccer for UConn, that was always something that I wanted to do. But as you get older, you realize the chances on some of those things are very slim. Maybe I wasn’t good enough. And then as I kind of got a little bit older, I started getting more and more into writing. I found I was very good at that. So that’s kind of like when I went into college where a lot of my focus was and I love sports. So in college, I really, I focused on journalism and mass media and that sort of thing. That’s where my communications degree wide, and then getting out of college, I had a lot of friends who had started to look in the space of like journalism, TV, broadcasting, all of those sorts of things. And I started looking at the hours for that, the opportunity for that, newspaper being sort of like a dying forum for information. And I had connections, and they helped me get in the door, at least to a place like MassMutual, and that’s really where I’ve been ever since.

JAMES: Wow. And that’s so rare to see you know a real connection there in the beginning that may you found a calling. Again, you didn’t go to school for insurance, right? And most folks in the insurance business don’t. Now, there are more and more and more college programs for insurance, which I’m really excited about. And certainly, I’m on the board of a public university in Houston called Texas Southern. I’m trying to get an insurance program kicked off there, we all want to show people the light early, but most of us stumble into this. I did. And you did. What was it about the industry and MassMutual specifically that kept you hooked for a decade and a half?

DAN: MassMutual has been around for a long, long time, the building that I go into every day is, I don’t know how old, 100 years old, more. What kind of kept me around, I guess, would be there’s just so much and so much opportunity and different types of roles that you can take on. That you can know without sort of jumping to different places, different organizations you can do pretty much whatever you want there. So it’s like if you want to move into technology, if you want to move into marketing, if you want to move into sales, all those sorts of opportunities are all there for you, being a large company. And I love the people. The people there are awesome, I have people that I’ve known for 16 years, and still see. A lot will come and go. Sometimes they leave and then they come back. But those types of connections definitely keep you around where you have that familiarity and it’s a company that’s frankly grown quite a bit over the last 10 years or so. There’s achievements that they’ve made in the last decade that, frankly, when I got there, I didn’t think they could get to some of those sites. So it’s definitely been a company that even though it’s a legacy brand, it’s really been on the rise quite a bit over the last 10, 15 years.

JAMES: Yes, I’ve been a policyholder MassMutual for a very long time, so a very consistent company. I like stability in my insurance companies. I do want innovation. I was talking about this, I was a speaker at the Insurance Council of Texas yesterday and as innovative as I like to be as innovative as I like my company and our products to be. As a customer of insurance, I value stability more than anything else. I want stability in my premium payments, I want stability in my dividend checks, I want you to be there when I have an issue. It’s something, I think, that we’ve got to always balance when we’re talking about insurtech. It just needs to deliver, technology needs to deliver stability as well as you know an improved user experience and customer experience and better improved financials.

So, you did a whole bunch of different jobs in this. And I think, obviously, the one that we’re talking about today, you’re a strategic business consultant internally. So you’re really an internal business consultant to the other sections of the company. Rather than them hiring a consultant externally, they have you going around and consulting with different organizations. You’ve really risen through the ranks. Tell me what you do now, and then we’re going to jump into RPA.

DAN: So, just to kind of make the connection, I talked about in school where I really focused was, I had a journalism minor when I was in school and that’s where I started. Through the years, and when I started at MassMutual I did a lot of writing, a lot of technical writing training that sort of stuff. But it really gave me a perspective of sort of like the whole company because we worked with all different areas within it, and I did that for a number of years, managed along the way as well. And about almost five years ago now, I sort of made the switch to get sort of in the business as opposed to supporting the business. So I moved to our claims organization with a primary focus as a consultant for robotic process automation, RPA. But once I got there, once you’re in the business, once you’re sort of tight to that, there’s all sorts of other opportunities that present themselves in addition to what they brought you there for. So my job right now is a lot of one ideation around sort of what are the things that can get us better, managing projects that are within the claims space, working with other areas outside of claims like marketing, and other areas that support the products that we serve. So it really runs the gamut, but a lot of the focus tends to be around how do we make that connection between our technology group and our frontline. I’ve sort of had this reputation of being able to speak business to business people, and speak technology to technology people, and be able to bridge those two worlds. So that’s really, I think, if I had to sum up, like kind of what I do, I think that captures it.

JAMES: It really is a unique skill to be able to communicate across that divide. The technical chasm and being able to understand what’s going on in technology enough to relay the message correctly and understand business well enough to understand the problem. And obviously the ideal state is that the technologists that are developing solution sets, understand the problem in the business, but that’s actually quite difficult because if they’re going to be effective technologists, it’s hard to have them spend enough time understanding the business. And so you’ve at some point, people have to specialize and then you need a bridge gapper, right? And that’s really just such a critical role. It’s also really good that you went to serving claims. That’s really neat. And why don’t you, since this is going to be our topic to jorn, then I’ll hand it off to Rob, just define what robotic process automation is and what kind of projects you’re working on briefly, and then I’ll let Rob jump into it.

DAN: I would say especially being at an organization that’s been around as long as MassMutual, and this is probably the case with a ton of insurers across the world, right? We have policies that have been on the books for 50, 60 years sometimes. And with that comes a lot of old systems and set processes that have been around forever. So, from an RPA perspective, really what we’re doing is we’re taking a lot of those existing processes that truly have remained stable over a number of years, and using technology allowing these, quote unquote, bots or automations right to run through the steps that a person would normally do. I would say one thing from that perspective where we had kind of started with it, versus where we’ve gone with it, we really started from this perspective of what can we automate. So looking at something and say, is that something that can be automated? Is this thing that this person does something that we can hand over to a bot? We learned that was difficult. And the reason why was because the way people do things and the way somebody might need to program one of these bots may be slightly different. People can use their eyes to do things which bots are now starting to be able to do a little bit of. People may be using things off the side of their desk or referencing things that a bot wouldn’t necessarily have the capability to do, and that’s where we sort of transitioned from that, what can we automate to more of a how can we automate? Taking a look at a process that maybe a person has done for you know the last 10 years, the same exact way, and looking at it and figuring out what are the little pieces of that of that process that presented a challenge for us from an automation perspective, and can we then tweak those things to make it work? Or are they even things that need to happen? Sometimes in organizations like this, especially if you have folks who have been doing the same job for a long time, you find things all the time that people have no idea why they’re doing them. They’ve done them a certain way for a long time. And you ask them, why do you do that? And they go, I don’t know if somebody told me to do that a long time ago. So sometimes along the way, we even throw away steps that just add no value to the process. But that’s really been the journey for us, and it was really a bumpy start for us, I would say. We didn’t really start off on the right foot with robotics, it took us a number of years to kind of catch our footing with it and really understand what it meant for us to do the process automation stuff. But we’ve gotten to a point now where we’ve automated a lot of the major tasks that we do, and we have a really good pipeline for how to continue to do that.

JAMES: That’s great. Rob?

ROB: Dan, I’d love to have you expand a little bit on some of the best practices that you’ve identified. I’ve led teams in the past where you know if you see robotic process automation in action, it can be… this miraculous thing. I think it was like a macro and steroids room back in the day. But I’ve also seen challenges where exactly to your point, you sit with somebody and you say, well, they do this task. It takes them 15 minutes and they do it 2,000 times a year or whatever it is. But things can change even in relatively stable processes. I remember one in the past where we were like logging into a department of insurance site or some type of thing to get some compliance information. All of a sudden, the government, the state government change their website, right? And all of a sudden none of the RPAs work, everything’s crashing, whatever. So anyway, I’m just kind of curious, maybe you can share some of those hard-worn learnings like that as you’ve established, like, now here’s some things that we found out that established this best practice.

DAN: One of the big things I think for that ah early on, they were doing a lot of sort of like using the UI, the user interface right to to actually like do the robotics, right? So they’d look at the way some, like going back to what can we automate. They would look at sort of what the person was doing and say, okay, let’s have a bot run through those steps. Walk through the screens the exact same way that person would walk through them, all of that. The problem with that is anytime anything changes, now we got to fix something, we got to update. And what we learned to do through the years is more and more try to use APIs and sort of go through the back end of these systems. So as opposed to hitting sort of the front end, which could change slightly, and in some cases, if it was like an external website, we wouldn’t even know about it. Those things that can change at a moment’s notice, take some of that out of it. Sometimes those APIs still have to change over time, but less likely. There’s more of a base there that doesn’t generally change as much. And that’s one thing that we really have tried to do is whenever possible, we’re not using sort of the UI. So the way somebody might do it, you know we’re going to hit all the same notes, but we’re doing it a different way.

JAMES: And it’s really important to sometimes lift ourselves out of this specific solution, and say, okay, let’s stop talking about like this one piece of technology, like we’re not a UiPath shop, right? Like UiPath is one tool, we might use Blue Prism, might use UiPath, we might use something completely different or something completely new, or that might not be the solution at all, maybe the real problem. And I and this is an often misattributed quote, it’s often misattributed to Einstein. He didn’t actually say it, but whoever said it said, if you have an hour to study, an hour to solve a problem, spend 55 minutes studying the problem and 5 minutes studying the solution. And we often square peg round hole solutions on top of problems saying, hey, you have a problem. Let’s slam, we already have a UiPath practice and team. Let’s slam UiPath on top of it. Automate that task instead of saying, hey, is there an API for that provider? Can we do an API crosswalk with ours and then you know have something that’s fully automated? I mean, that would be really great. And of course, the nice thing about APIs is they’re intended to be accessible machines. So they tend to be in change control processes and version control. And you can have some things deprecated out of an API feed. But it’s very, it’s managed, whereas, RPA can sometimes be mass chaos, right? Like UiPath has the ability to adapt to a screen that has minor changes, but at some point, you have to rewrite the bot.

DAN: For sure. When we write about the first time, we know that that book is not closed. We know that at some point, we’re going to have to do updates. We know that at some point, we’re going to want to enhance it. So when we try to look at those opportunities, we try to get as much as possible that can be done in an automated fashion that’s not problematic. But we also see as the person who is sort of the go-to for our claims organization for this type of work, I get a lot of things thrown at me where somebody will say, hey, like why don’t we have a bot do this? Sometimes they’re too small. Sometimes they’re the not not the right type of use case for robotics. And what I try to do is really focus on the problem as opposed to could this be a bot? So that’s part of my job as a strategic consultant is trying to help folks think through some of those things, whether it’s different reporting ah that we can provide somebody, whether it is just maybe a simple process change or some way to stream data that we hadn’t thought of before. All of those things sort of come into play. Like I said, I’m known sort of for the robotic space, but when that problem comes across my desk, it’s really about okay, what are we really trying to solve here, and are you just trying to get something annoying to go away, right? That’s something I find a lot. A lot of times what I see is somebody will come to me and they’ll say, hey, it would be great if we had a bot for this. And then I start to ask more questions and figure out how often does this happen? Oh, it happens once a week. It’s like, that’s probably not going to work. It’s not something that we would build for. But what can we do? What are the options we have at our disposal? That’s the way I always try to approach it, it’s like I write everything down that comes to me and nothing is sort of thrown in the trash. But there’s a right place for robotics and there’s a wrong place for it, and we’ve got to build in the right ways.

JAMES: And when we mean, you know, robotics, of course, we’re talking about software robots, not hardware robots. Just with just for the folks out there that tend to think that robots are only hardware, these are software robots that, the best equivalent for those of you who might be having a difficult time with this is macros in Excel, where you can record something, except UiPath and Blue Prism, or tools like that can adapt to the screen. So they actually have the ability to learn and adapt, whereas if you change one cell on your Excel, on your VB macro, it’s going to break the whole thing in Excel. So just keep that in mind too. I also like that you’re calling out probably the most important question anybody in any innovation or technology organization that any insurance company should ask. And you called it out. You said, what’s the real problem here? I always feel like people should just print that out and put it on the door because people will tell you what solution they want, but that solution is constrained by the box that they live in. So if you really dive into it, we had a recent issue with claims as well, where we were looking at EOBs being attached to the payments and then we were being told by the adjuster that what they needed was this very specific thing. I need the EOB and here I need to be able to print out both and then attach them later. We’re like, what if we just put it all together and bundled it up front for you? So you didn’t have to do that. What if we pre bundled it and we made it seamless and we automated the entire process rather than giving you the output that you’re asking for, which is the output your 20 year old system had? Let’s rethink EOBs being attached to payments and how we’re going to do it.

DAN: Sometimes they’re imagining like the simplest solution that would get it solved for them, right? And not thinking about maybe like what the real issue is, how to solve it.

JAMES: Yeah. Rob?

ROB: Actually, that’s exactly what I was going to follow up, Dan. So obviously, you know you don’t have a claims background. I’m sure many of the people that you do work with do. They’ve you know lived claims day in, day out. And one of the struggles that I see is that um people that work in particular systems, they have a hard time thinking about process independent of the systems, right? The system is the process. And so I’m imagining you come in and you have that, even though you know you’re inside MassMutual, you’re kind of coming in with claims knowledge, but not claims expertise. And so that gives you, you probably know enough just to be dangerous, right? But you kind of are bringing a little bit of that outsider’s perspective. So we’d just love for you to maybe share a little bit about, particularly in claims, like other certain processes, whether it’s first know loss, subrogation, you know, getting medical records, like, is there any, but use cases, that you’ve seen kind of work really well? And then conversely, maybe there’s a few that really robotic process automation may not be the right solution for based on what you’ve seen.

DAN: I would say, some of the things that, the way I would have answered this question probably three years ago is probably different from the way I would answer it today. And what I mean by that is technology grows so much. So a lot of the use cases that we would have thrown out three, four years ago are now on the table again. From sort of a claims perspective, some of the things that we’ve seen that have been really successful in automating, sending out letters is a big one. Or follow-up emails, those sorts of things. Automated text messaging is something that we’ve done a lot recently. You actually mentioned medical records, we actually have some automation in our DI claims space where we do ordering of medical records. And that one’s actually a perfect example, we talked about APIs. That bot that we had was originally built off of the user interface, and every single quarter, we’d have to double check to see if it worked still. About a year and a half ago or so, it got rewritten where now it was going in through the API that was created for it. So those are just like those are examples of where you know even in three, four years’ time where stuff changes.

Another good example of a business case where we would have thrown it out, the ability to do text recognition or character recognition, the ability to read forms. When I started doing robotics, the idea that you could take a form that came in, have a bot read it, determine what’s on it, and then do something with it was basically…

JAMES: Wizardry.

DAN: Wizardry. Right. And now we have a few bots that do that work. There’s some manual intervention when somebody’s got sloppy handwriting or something. In those cases where somebody actually like hand wrote and filled out a form when it comes back in. For the most part, we can program it. So that it knows what fields it’s looking for, it knows how to grab that information and then actually do something with it. So that’s why I say like a lot of the business cases we didn’t we didn’t even think were on the table are now on the table. So it was like, I think we got to a point probably two years ago where some of the bigger blocks of work that we thought were the RPA cases that we were going to get. We got to those in a lot of ways. We did those. And then as technologies sort of move forward and we’ve had more opportunity to to use some of it, we’ve seen just new business cases be able to spring up. So it’s something that at some point, you know I always say with a lot of these automations that we do, ultimately, we’d like to get rid of them, because some of this stuff is built off of old systems that we hope go away. Or, you know, I talk about reading information off of forms. It’d be great if I didn’t even get a form sent to me. So, like, in a lot of ways, it’s sort of like patching up sort of legacy software or and administrative systems that we have. And we hope that someday some of this stuff goes away. But as you guys probably know, some of that stuff is not fast moving in the insurance space. So getting rid of old systems that may have 50-year-old policies on it is not exactly something that goes quickly. So those things we know will be around for a while. So it’s a matter of you know continuing to up the ante and move to the next spot where we think we can be useful with this stuff.

JAMES: It’s a moving target though. I mean, look, the RPA capabilities are really dramatically expanding just the – and especially with once you integrate RPA because robotic process automation, the ability for a software robot to read a screen and read anything on the screen and then take action based on that. That’s been around long before large language models really were mass mass popularized. Well, now I can summarize the content of the screen, tell you about it, and actually draft a response for you. I mean, the capabilities are dramatically expanded now with, and so that does give you a few more weapons in the toolset. Ultimately, though, I kind of feel like a lot of these technologies are transitional technologies, that are meant to get us towards. Ideally, a fully integrated future, if you have a fully integrated system, there’s a lot there’s a lot of things you don’t need, right?

DAN: Yeah, the more people sort of like interact with us digitally, the more we have to sort of do a lot of this stuff, the more we can build a straight through processing and do that sort of automation as opposed to the robotic process automation.

JAMES: Yeah, exactly. Very cool. Rob, you want to bring us home?

ROB: Dan, just talking about and curious about your thoughts. In the past, I know use cases are articulated to me that were challenging a few years ago. You might identify the word surgery, for instance, right as part of handwritten notes from a doctor. But what you didn’t have is a contextualization to say no surgery is required, or surgery may be required if physical therapy is not successful, or surgery is required immediately and that may require different actions by a claims adjuster that might be routed to different adjusters based on their expertise and and things like that. And so those were some of the more demanding I guess use cases four or five years ago. And it sounds like a lot of that now, through combining with AI, large language models are possible today. So I just kind of curious, like where do you see the future going with some of these technologies going forward.

DAN: So that stuff, when I originally came to claims, one of the big things they were looking at was sort of natural language processing. How do you take text that’s on a page, whether it be handwritten or typed, make sense out of it, and then be able to use AI to make decisions off of it. And, you know, four years ago, there was not a lot of like well established solutions for this stuff that was readily available like it was starting, and we were sort of like knocking on that door but we never really felt like there was anything out there at the time at least right that could hit all the all the aspects we wanted to hit. To your point, if you’re using those sorts of processing models to pull up any sort of useful information using AI, you’ve got to be able to siphon out the stuff that’s like in the way. One thing we noticed with certain things was if we had a natural language and processor looking at, like medical records, and we’d look for a date of a surgery. A lot of the models would pick up the date that the form was sent over. A lot of that, I think, has continued to evolve. We’re actually kind of moving back into that space after sort of taking a break from it um and trying to use AI models to sort of better understand, especially like on the DI side where but we’re sort of managing these blocks of claims as opposed to with life and annuity death claims, which we do, which are sort of sort of like in the door, out the door. But for disability claims, it’s how do we manage this block and understand where are the opportunities to follow up with folks to potentially close claims. And we’re going to start over the next year looking at that more. And we have some solutions at our disposal now that we think can be a great help for that because that world is extremely complicated. Trying to deal with folks who are on disability, the different, something like back pain, right? Back pain. How that could be, there’s the range of sort of limitations somebody could have from something like that is huge. So being able to use AI or natural language processing models to be able to cut through some of that, and give us sort of a sense of like where to go, with something like that, um we think is going to become a big part of how we operate.

JAMES: Very cool. Well, this is an excellent conversation. I actually would like a lot of our listeners to listen to it because there’s so many practical takeaways. I want to point out that In this discussion, we are discussing the use of AI as one toolset, but not as “the” tool. In fact, we actually don’t need for, you know, we have 280 employees, we work for insurance companies, right? We build software. What we do, we build software for insurance companies. We can solve most of our clients’ problems without using AI. Like just in general, the most of the issues, if you look at if, you look at the average backlog for the average insurance company of things they need and problems they have, most of them can be solved without ever leveraging AI at all. So AI is not the end all be all. It’s not the multi-tool that solves everything. It is an amazing tool. It’s incredible. And I really enjoy it. But, what you’re hearing is that there’s a whole lot of meat and potato solutions that can be solved very directly. And that’s exciting, it’s exciting. I’m excited for you, especially with what you can do for MassMutual. So I want to wrap up with this final question. I’m a sci-fi and Star Trek fan. Are you Star Trek or Star Wars?

DAN: I would be Star Wars for sure. I’ve seen some Star Trek, but not much.

JAMES: Ok, I can live with that here. I just need you to remember that Star Wars is historical fiction because it was a long time ago in a galaxy far, far away. So it’s past tense and Star Trek is the future. So it’s one. Yeah. So no one’s fantasy, one’s sci-fi. That’s OK. It’s cool. I actually was not into Star Wars until of all things Mandalorian brought me across Mando. Mando brought me across. It’s a really good show. And I was like, man, this is really good. So that’s when I actually kind of got into it.

DAN: I don’t want to compare you to my kids, but that’s when my kids got into that too, because they saw that little baby Yoda, right?

JAMES: Dude, how can you not see? I have a stuffed baby Yoda at the house. Like, how can you not see baby Yoda and want not just love you’re like, oh, I do know baby Yoda.

DAN: The funny thing is whenever we would watch that show, you know there’s all these gaps where Baby Yoda is not really pictured. And they’d be like, can you please fast forward because this is not doing it to me.

JAMES: We just want to watch Baby Yoda.

DAN: Yeah, that’s it. Yeah, just show us Baby Yoda.

JAMES: So thinking in the vein of Star Trek and looking to the future, like what big leap do you feel like is coming up or that you’ve seen that has yet to be leveraged?

DAN: The AI thing has become so prevalent. So like you said, it’s not the tool for us to solve all problems, but we’re seeing that it’s going to be increasingly one of the major tools that we leverage. I think for a company like MassMutual, where we have such a long history, a lot of our goals center around like just how do we deal with people more digitally? How do we get people online that may have had a policy for so long that they forgot about it? Those sorts of things are huge for us. So we’ve really tried to focus a lot on building out that digital experience from new business all the way to claim. And not just for our customers, but for our advisors. The days of doing business on paper, you know we still get a lot of paper. But how do we like move the needle and start to move away from that? Because really what it enables us to do at the end of the day is get ourselves some data, actually, right as opposed to stuff that’s just written down. So I think that’s one of the huge pushes and that really, for us, I think will sort of be the start of how do we then leverage things like AI to then educate our business and what we need to do. Because, you know, and you know I talked to some folks, who I’ve been with our claims department for for five years. Some of the folks that are in our department have been around there for 20 plus. And they’ll tell me about the days where when somebody filed the claim, they had to send a request to get the literal paper that somebody submitted the policy on, mailed like interoffice mail to them. And then, and then yeah.

JAMES: Interoffice mail, the good old days.

DAN: And then go to, put that in a binder and start to collect all the paperwork for that. We’d require sort of like wet signatures, all that stuff. There’s people that I know of in our claims department who actually work with typewriters at one point. So it’s like, we’re coming from a world where like that is still in some people’s minds of that time. But the way things evolve, the leaps become so much quicker. When you look at something like a chat GPT and how quickly that got to a space, where you had a million people using it…

JAMES: 100 million. It was like, it was super fast to 100 million.

DAN: So it’s like the changes, the rate of change has continued to just increase and increase and increase. And I think as a company, MassMutual has always been in sort of like a very good standing. We kind of started this by talking about stability. And they’ve been around for a very long time. But I think at the same time, like, we’ve tried to do a good job in staying up with technology because, there’s insurance companies that haven’t or have been unstable and have gone away. So it’s not an unheard of thing even for something, a company that’s been around as long as MassMutual. So we’re always looking towards the future and what those next things are. And I know from for for a fact that there’s already things in the works for us in terms of like working with new AI models and how do we expand our digital base and all that for our customers.

JAMES: Exactly. Awesome. Well, thank you for your time today. And thanks for talking with us about all these cool ways you’re implementing tech inside MassMutual.

DAN: Absolutely.

JAMES: And appreciate it. Thanks so much. And Rob, as always, thank you for joining in.

ROB: Yeah, great episode. Thanks for the conversation, Dan.

JAMES: To our listeners, thank you for tuning in today to geek out our interview with Dan Savage from MassMutual. See you next time. Enjoy the ride and geek out.